| Posted | Nick | Remark | |
|---|---|---|---|
| #openstack-nova - 2017-03-08 | |||
| 17:42:52 | mriedem | compute agent? | |
| 17:42:59 | mriedem | we don't have a compute agent | |
| 17:43:19 | sean-k-mooney | n-cpu? | |
| 17:43:46 | cfriesen | compute server | |
| 17:43:48 | sfinucan | mriedem, sean-k-mooney: Might be easier show you the code :) https://review.openstack.org/#/c/379625/10 | |
| 17:44:05 | cfriesen | make that compute service | |
| 17:44:16 | mriedem | sfinucan: ok that's not the compute driver | |
| 17:45:05 | sean-k-mooney | mriedem: right what sfinucan is proposing to change is how the pci device is selected when it lands on the compute node but i tink that should be driver independent | |
| 17:45:08 | mriedem | so the compute_nodes.numa_topology field has the data you need, which we get in the resource tracker | |
| 17:45:36 | mriedem | sean-k-mooney: i thought the point was to not land instances that don't require pci on hosts that have available pci inventory? | |
| 17:45:58 | mriedem | doing that calculation after we've picked the host is too late | |
| 17:46:13 | sean-k-mooney | mriedem: that is where the weigher comes in | |
| 17:46:36 | sean-k-mooney | mriedem: untill we have claims you need to do it twice | |
| 17:46:45 | sfinucan | mriedem: Yeah, it's two steps | |
| 17:47:00 | sfinucan | Step A: Don't use hosts with PCI devices if we don't need them (the weigher) | |
| 17:47:29 | sean-k-mooney | if you run out of server that dont have pci decices and you have to land on one that does then you should still prefer a numa node that does not have pci devices attached | |
| 17:47:35 | sfinucan | Step B: If we do get a host with PCI devices, avoid NUMA nodes with PCI devices if we don't need them | |
| 17:47:59 | sean-k-mooney | yep what sfinucan said :) | |
| 17:50:47 | mriedem | i'll be honest, my eyes glaze over once i see numa_fit_instance_to_host | |
| 17:53:39 | sean-k-mooney | mriedem: that is fair. if i didnt have to know about numa effect for my job i would happy delete it form my mind | |
| 17:55:48 | sean-k-mooney | sfinucan: so stepen to sergeys vis versa comment are you going to extend the weigher to also perfer numa nodes with pci devices available when they are requested or just update the comment ? | |
| 17:56:54 | sfinucan | sean-k-mooney: I think I'll just remove the comment and rely on balancing with CPUs | |
| 17:57:31 | sean-k-mooney | sfinucan: ok its a 30 second else clause with the same labmda but the order reversed to add if you wanted too | |
| 17:58:11 | sfinucan | sean-k-mooney: But it's another 10 minutes writing unit tests ;) | |
| 17:58:13 | sean-k-mooney | sfinucan: i havent thaugh however if it is desireable | |
| 17:59:01 | sean-k-mooney | haha ya i was impressed that the ratio of test lines to code was actull 29:11 | |
| 17:59:50 | sean-k-mooney | less then 4 time the amount of code in unit tests then busness logic is rare | |
| 18:02:34 | openstackgerrit | Eric Fried proposed openstack/nova master: PowerVM Driver: spawn/destroy #2: functional https://review.openstack.org/438598 | |
| 18:02:39 | inc0 | good morning people, http://lists.openstack.org/pipermail/openstack-dev/2017-March/113507.html <- this issue might interest you | |
| 18:03:18 | inc0 | pretty much libvirt 2.5 seems to be broken with current nova vm datafiles creation | |
| 18:05:09 | inc0 | I'm not sure if it's ubuntu-specific or it's libvirt 2.5 specific, but it is broken:) | |
| 18:05:31 | inc0 | and we have few alternatives to handle this | |
| 18:05:56 | inc0 | one would be to allow nova to create these files with custom umask | |
| 18:06:13 | inc0 | today it's hard 722, if we could make it 772, that would do it | |
| 18:06:27 | inc0 | umask and ownership | |
| 18:07:44 | sean-k-mooney | inc0: the umask is the revers of what you think it is. where 777 is read write execute normally for umask its 000 | |
| 18:08:40 | inc0 | fair enough;) | |
| 18:08:41 | sean-k-mooney | inc0: you would need umask 002 | |
| 18:08:45 | inc0 | well, file mode | |
| 18:11:20 | sean-k-mooney | inc0: i have a vm using devstack with 2.5 currently and i can boot vms. i know on the kolla channel tere was an issue with the console file correct | |
| 18:11:35 | sean-k-mooney | is that what your refering too or is more then just the console | |
| 18:11:45 | openstackgerrit | Stephen Finucane proposed openstack/nova-specs master: Reserve NUMA nodes with PCI devices attached https://review.openstack.org/364468 | |
| 18:14:01 | openstackgerrit | Eric Fried proposed openstack/nova master: PowerVM Driver: power_on/off and reboot https://review.openstack.org/427380 | |
| 18:14:01 | sfinucan | mriedem: I've clarified the fact that there's no real changes to the driver (the location of the code confused me and I wrote that code a good while back) | |
| 18:14:17 | openstackgerrit | Eric Fried proposed openstack/nova master: PowerVM Driver: spawn/destroy #3: TaskFlow https://review.openstack.org/438729 | |
| 18:16:38 | sfinucan | As for the resource tracker stuff: I'd be inclined to suggest that the change is both helpful/small enough, and nested resource providers still fuzzy enough, to warrant taking the weigher route for now | |
| 18:17:24 | sfinucan | However, I'm not sure what the plan is for weighers and what the migration path is when we move to resource providers, so I'd definitely like jaypipes' input | |
| 18:17:42 | sfinucan | For now though, it's late and the Lego Batman movie isn't going to see itself :) | |
| 18:18:16 | sean-k-mooney | sfinucan: nice i must go see that soon. | |
| 18:18:28 | sean-k-mooney | sfinucan: let me know if its any good :) | |
| 18:21:10 | bauzas | mriedem: I +W'd https://review.openstack.org/#/c/442723/1 but I think we could possibly have follow-ups | |
| 18:21:14 | bauzas | see commentqs | |
| 18:21:25 | openstackgerrit | Dan Peschman proposed openstack/nova master: No API cell up-call to delete consoleauth tokens https://review.openstack.org/415922 | |
| 18:21:58 | openstackgerrit | Dan Smith proposed openstack/nova master: Add functools.wraps to wrapper https://review.openstack.org/443175 | |
| 18:22:08 | mriedem | dansmith: no ice_t? | |
| 18:22:28 | dansmith | mriedem: if I can't have my fun you can't either | |
| 18:22:45 | mriedem | works for me | |
| 18:22:52 | bauzas | dansmith: wow, chilling at the series https://review.openstack.org/#/c/441461/8 | |
| 18:23:04 | dansmith | the first time it came up, I was going to rename it to gangsta_rapper | |
| 18:23:06 | sean-k-mooney | inc0: what is the actully error that ye are seeing with libvirt 2.5? | |
| 18:23:20 | dansmith | bauzas: all but one tempest test is passing with multiple cells with that series :) | |
| 18:23:33 | bauzas | \o/ | |
| 18:23:37 | bauzas | heh for https://review.openstack.org/#/c/443175/3/nova/conductor/manager.py | |
| 18:23:41 | dansmith | and I had that working before, but broke it in cleanup | |
| 18:23:46 | bauzas | that remembers me some comment I made :) | |
| 18:25:09 | bauzas | dansmith: any reason why you added https://review.openstack.org/#/c/443175/3/nova/conductor/manager.py as a top patch and not just modified https://review.openstack.org/#/c/438022/15 ? | |
| 18:25:21 | sean-k-mooney | inc0: oh appolgies my libvert is 1.3.1 for some reason i guess i dont have the cloud archive enabled | |
| 18:25:22 | dansmith | bauzas: yes, see the comments | |
| 18:25:26 | bauzas | oh ok | |
| 18:25:30 | openstackgerrit | Andreas Jaeger proposed openstack/nova master: Replace obsolete vanity openstack.org URLs https://review.openstack.org/443266 | |
| 18:25:32 | dansmith | bauzas: I would eat all infra re-pushing for that nit | |
| 18:26:20 | bauzas | dansmith: um ok | |
| 18:26:23 | mriedem | dansmith: comments inline https://review.openstack.org/#/c/438022/15 | |
| 18:29:36 | dansmith | mriedem: replies inline.. I'd tack the log fix into the rapper fix if it were me, but your call | |
| 18:30:09 | dansmith | if you want it fixed, I'll just squash that nit patch into this one | |
| 18:30:20 | dansmith | fixed pre-merge I mean | |
| 18:31:13 | mriedem | dansmith: question about the decorator on build_instances - that's not just cellsv1, it's also reschedules from the compute manager - but at that point we're already targeted right? | |
| 18:31:40 | dansmith | mriedem: right, and so we can't actually look up there, so the target would have to be graceful and fail, yet it'd be ugly | |
| 18:31:49 | dansmith | mriedem: that goes back to the cell conductor instead of super | |
| 18:32:13 | mriedem | oh yeah, need that diagram | |
| 18:32:44 | mriedem | so i'm ok with fixing the log thing in the rapper patch, but can you also put a comment in that patch for both the build_instances thing and mdbooth's question that had a long explanation? | |
| 18:32:55 | dansmith | yep | |
| 18:33:00 | mriedem | word homey | |
| 18:33:31 | mriedem | also, | |
| 18:33:38 | mriedem | no one mentioned your flock of seagulls reference | |
| 18:33:41 | mriedem | pick a genre | |
| 18:33:52 | dansmith | mriedem: I know, I thought that was amusing :) | |
| 18:34:00 | mriedem | i got it though | |
| 18:34:59 | openstackgerrit | Eric Fried proposed openstack/nova master: PowerVM Driver: spawn/destroy #4: full flavor https://review.openstack.org/391288 | |
| 18:35:14 | openstackgerrit | Eric Fried proposed openstack/nova master: PowerVM Driver: console https://review.openstack.org/409402 | |
| 18:35:50 | openstackgerrit | Dan Smith proposed openstack/nova master: Address minor feedback in superconductor targeting https://review.openstack.org/443175 | |
| 18:36:17 | dansmith | mriedem: at least I was decade consistent | |
| 18:36:23 | dansmith | afaik, nobody calls them "rappers" anymore | |
| 18:36:32 | dansmith | but they did around the seagulls timeframe | |
| 18:36:32 | mriedem | what are they called now? | |
| 18:36:54 | dansmith | um, I dunno, "Hip Hop Artists" ? | |
| 18:37:12 | mriedem | yeah maybe | |
| 18:37:14 | mriedem | gdi kids | |
| 18:37:16 | mriedem | *gd | |
| 18:37:19 | dansmith | I feel like if someone said "who's your favorite rapper" nowadays you'd get laughed at, but in the 80s that was a serious question | |